This laboratory has developed the first reliable and convenient animal model for the very distinct group of lesions characteristic of human background diabetic retinopathy. Taking advantage of the fact that aldose reductase has a higher affinity for galactose than for glucose, which results in more intracellular polyol accumulation in galactosemia, we fed rats galactose for extended periods in an effort to produce diabetic-like retinal lesions. Galactosemia indeed induced diabetic-like microangiopathies that were more advanced and more like human diabetic lesions than those which develop in long-term diabetic rats. The galactose-fed rat exhibits capillary basement membrane thickening within 28 weeks; selective pericyte loss, capillary dilation, endothelial cell proliferation, some tortuosity of vessels, and some acellularity within 33 weeks; microaneurysms, occlusions, and shunts within 66 weeks; and extensive regions of varicose capillary meshwork by 98 weeks, making the galactosemic rat a good model for diabetes-induced retinal microangiopathies that take longer to develop. This rat should also serve as a model for other polyol-related complications of diabetes. The galactose-fed rat model has distinct advantages over genetic or chemically induced models of diabetes for intervention studies: It shows lesions sooner, and upon removal from the galactose diet, it returns to a normal physiological state within a few days. We plan to determine appropriate times for intervention using different aldose reductase inhibitors, and to attempt, by dietary manipulation, to produce rat models that develop the diabetic-like retinal angiopathies sooner. Also, using cell culture, we will investigate possible mechanisms of endothelial cell proliferation and subsequent pathologies.